Explainable deep learning-based survival prediction for non-small cell lung cancer patients undergoing radical radiotherapy.

BACKGROUND AND PURPOSE: Survival is frequently assessed using Cox proportional hazards (CPH) regression; however, CPH may be too simplistic as it assumes a linear relationship between covariables and the outcome. Alternative, non-linear machine learning (ML)-based approaches, such as random survival forests (RSFs) and, more recently, deep learning (DL) have been proposed; however, these techniques are largely black-box in nature, limiting explainability. We compared CPH, RSF and DL to predict overall survival (OS) of non-small cell lung cancer (NSCLC) patients receiving radiotherapy using pre-treatment covariables. We employed explainable techniques to provide insights into the contribution of each covariable on OS prediction. MATERIALS AND METHODS: The dataset contained 471 stage I-IV NSCLC patients treated with radiotherapy. We built CPH, RSF and DL OS prediction models using several baseline covariable combinations. 10-fold Monte-Carlo cross-validation was employed with a split of 70%:10%:20% for training, validation and testing, respectively. We primarily evaluated performance using the concordance index (C-index) and integrated Brier score (IBS). Local interpretable model-agnostic explanation (LIME) values, adapted for use in survival analysis, were computed for each model. RESULTS: The DL method exhibited a significantly improved C-index of 0.670 compared to the CPH and a significantly improved IBS of 0.121 compared to the CPH and RSF approaches. LIME values suggested that, for the DL method, the three most important covariables in OS prediction were stage, administration of chemotherapy and oesophageal mean radiation dose. CONCLUSION: We show that, using pre-treatment covariables, a DL approach demonstrates superior performance over CPH and RSF for OS prediction and use explainable techniques to provide transparency and interpretability.

Voxel-Wise Comparison of Co-Registered Quantitative CT and Hyperpolarised Gas Diffusion-Weighted MRI Measurements in IPF.

The patterns of idiopathic pulmonary fibrosis (IPF) lung disease that directly correspond to elevated hyperpolarised gas diffusion-weighted (DW) MRI metrics are currently unknown. This study aims to develop a spatial co-registration framework for a voxel-wise comparison of hyperpolarised gas DW-MRI and CALIPER quantitative CT patterns. Sixteen IPF patients underwent 3He DW-MRI and CT at baseline, and eleven patients had a 1-year follow-up DW-MRI. Six healthy volunteers underwent 129Xe DW-MRI at baseline only. Moreover, 3He DW-MRI was indirectly co-registered to CT via spatially aligned 3He ventilation and structural 1H MRI. A voxel-wise comparison of the overlapping 3He apparent diffusion coefficient (ADC) and mean acinar dimension (LmD) maps with CALIPER CT patterns was performed at baseline and after 1 year. The abnormal lung percentage classified with the LmD value, based on a healthy volunteer 129Xe LmD, and CALIPER was compared with a Bland-Altman analysis. The largest DW-MRI metrics were found in the regions classified as honeycombing, and longitudinal DW-MRI changes were observed in the baseline-classified reticular changes and ground-glass opacities regions. A mean bias of -15.3% (95% interval -56.8% to 26.2%) towards CALIPER was observed for the abnormal lung percentage. This suggests DW-MRI may detect microstructural changes in areas of the lung that are determined visibly and quantitatively normal by CT.

PhysVENeT: a physiologically-informed deep learning-based framework for the synthesis of 3D hyperpolarized gas MRI ventilation.

Functional lung imaging modalities such as hyperpolarized gas MRI ventilation enable visualization and quantification of regional lung ventilation; however, these techniques require specialized equipment and exogenous contrast, limiting clinical adoption. Physiologically-informed techniques to map proton (1H)-MRI ventilation have been proposed. These approaches have demonstrated moderate correlation with hyperpolarized gas MRI. Recently, deep learning (DL) has been used for image synthesis applications, including functional lung image synthesis. Here, we propose a 3D multi-channel convolutional neural network that employs physiologically-informed ventilation mapping and multi-inflation structural 1H-MRI to synthesize 3D ventilation surrogates (PhysVENeT). The dataset comprised paired inspiratory and expiratory 1H-MRI scans and corresponding hyperpolarized gas MRI scans from 170 participants with various pulmonary pathologies. We performed fivefold cross-validation on 150 of these participants and used 20 participants with a previously unseen pathology (post COVID-19) for external validation. Synthetic ventilation surrogates were evaluated using voxel-wise correlation and structural similarity metrics; the proposed PhysVENeT framework significantly outperformed conventional 1H-MRI ventilation mapping and other DL approaches which did not utilize structural imaging and ventilation mapping. PhysVENeT can accurately reflect ventilation defects and exhibits minimal overfitting on external validation data compared to DL approaches that do not integrate physiologically-informed mapping.

A hybrid model- and deep learning-based framework for functional lung image synthesis from multi-inflation CT and hyperpolarized gas MRI.

BACKGROUND: Hyperpolarized gas MRI is a functional lung imaging modality capable of visualizing regional lung ventilation with exceptional detail within a single breath. However, this modality requires specialized equipment and exogenous contrast, which limits widespread clinical adoption. CT ventilation imaging employs various metrics to model regional ventilation from non-contrast CT scans acquired at multiple inflation levels and has demonstrated moderate spatial correlation with hyperpolarized gas MRI. Recently, deep learning (DL)-based methods, utilizing convolutional neural networks (CNNs), have been leveraged for image synthesis applications. Hybrid approaches integrating computational modeling and data-driven methods have been utilized in cases where datasets are limited with the added benefit of maintaining physiological plausibility. PURPOSE: To develop and evaluate a multi-channel DL-based method that combines modeling and data-driven approaches to synthesize hyperpolarized gas MRI lung ventilation scans from multi-inflation, non-contrast CT and quantitatively compare these synthetic ventilation scans to conventional CT ventilation modeling. METHODS: In this study, we propose a hybrid DL configuration that integrates model- and data-driven methods to synthesize hyperpolarized gas MRI lung ventilation scans from a combination of non-contrast, multi-inflation CT and CT ventilation modeling. We used a diverse dataset comprising paired inspiratory and expiratory CT and helium-3 hyperpolarized gas MRI for 47 participants with a range of pulmonary pathologies. We performed six-fold cross-validation on the dataset and evaluated the spatial correlation between the synthetic ventilation and real hyperpolarized gas MRI scans; the proposed hybrid framework was compared to conventional CT ventilation modeling and other non-hybrid DL configurations. Synthetic ventilation scans were evaluated using voxel-wise evaluation metrics such as Spearman's correlation and mean square error (MSE), in addition to clinical biomarkers of lung function such as the ventilated lung percentage (VLP). Furthermore, regional localization of ventilated and defect lung regions was assessed via the Dice similarity coefficient (DSC). RESULTS: We showed that the proposed hybrid framework is capable of accurately replicating ventilation defects seen in the real hyperpolarized gas MRI scans, achieving a voxel-wise Spearman's correlation of 0.57 ± 0.17 and an MSE of 0.017 ± 0.01. The hybrid framework significantly outperformed CT ventilation modeling alone and all other DL configurations using Spearman's correlation. The proposed framework was capable of generating clinically relevant metrics such as the VLP without manual intervention, resulting in a Bland-Altman bias of 3.04%, significantly outperforming CT ventilation modeling. Relative to CT ventilation modeling, the hybrid framework yielded significantly more accurate delineations of ventilated and defect lung regions, achieving a DSC of 0.95 and 0.48 for ventilated and defect regions, respectively. CONCLUSION: The ability to generate realistic synthetic ventilation scans from CT has implications for several clinical applications, including functional lung avoidance radiotherapy and treatment response mapping. CT is an integral part of almost every clinical lung imaging workflow and hence is readily available for most patients; therefore, synthetic ventilation from non-contrast CT can provide patients with wider access to ventilation imaging worldwide.

Implementable Deep Learning for Multi-sequence Proton MRI Lung Segmentation: A Multi-center, Multi-vendor, and Multi-disease Study.

BACKGROUND: Recently, deep learning via convolutional neural networks (CNNs) has largely superseded conventional methods for proton (1 H)-MRI lung segmentation. However, previous deep learning studies have utilized single-center data and limited acquisition parameters. PURPOSE: Develop a generalizable CNN for lung segmentation in 1 H-MRI, robust to pathology, acquisition protocol, vendor, and center. STUDY TYPE: Retrospective. POPULATION: A total of 809 1 H-MRI scans from 258 participants with various pulmonary pathologies (median age (range): 57 (6-85); 42% females) and 31 healthy participants (median age (range): 34 (23-76); 34% females) that were split into training (593 scans (74%); 157 participants (55%)), testing (50 scans (6%); 50 participants (17%)) and external validation (164 scans (20%); 82 participants (28%)) sets. FIELD STRENGTH/SEQUENCE: 1.5-T and 3-T/3D spoiled-gradient recalled and ultrashort echo-time 1 H-MRI. ASSESSMENT: 2D and 3D CNNs, trained on single-center, multi-sequence data, and the conventional spatial fuzzy c-means (SFCM) method were compared to manually delineated expert segmentations. Each method was validated on external data originating from several centers. Dice similarity coefficient (DSC), average boundary Hausdorff distance (Average HD), and relative error (XOR) metrics to assess segmentation performance. STATISTICAL TESTS: Kruskal-Wallis tests assessed significances of differences between acquisitions in the testing set. Friedman tests with post hoc multiple comparisons assessed differences between the 2D CNN, 3D CNN, and SFCM. Bland-Altman analyses assessed agreement with manually derived lung volumes. A P value of <0.05 was considered statistically significant. RESULTS: The 3D CNN significantly outperformed its 2D analog and SFCM, yielding a median (range) DSC of 0.961 (0.880-0.987), Average HD of 1.63 mm (0.65-5.45) and XOR of 0.079 (0.025-0.240) on the testing set and a DSC of 0.973 (0.866-0.987), Average HD of 1.11 mm (0.47-8.13) and XOR of 0.054 (0.026-0.255) on external validation data. DATA CONCLUSION: The 3D CNN generated accurate 1 H-MRI lung segmentations on a heterogenous dataset, demonstrating robustness to disease pathology, sequence, vendor, and center. EVIDENCE LEVEL: 4. TECHNICAL EFFICACY: Stage 1.

A Dual-Channel Deep Learning Approach for Lung Cavity Estimation From Hyperpolarized Gas and Proton MRI.

BACKGROUND: Hyperpolarized gas MRI can quantify regional lung ventilation via biomarkers, including the ventilation defect percentage (VDP). VDP is computed from segmentations derived from spatially co-registered functional hyperpolarized gas and structural proton (1 H)-MRI. Although acquired at similar lung inflation levels, they are frequently misaligned, requiring a lung cavity estimation (LCE). Recently, single-channel, mono-modal deep learning (DL)-based methods have shown promise for pulmonary image segmentation problems. Multichannel, multimodal approaches may outperform single-channel alternatives. PURPOSE: We hypothesized that a DL-based dual-channel approach, leveraging both 1 H-MRI and Xenon-129-MRI (129 Xe-MRI), can generate LCEs more accurately than single-channel alternatives. STUDY TYPE: Retrospective. POPULATION: A total of 480 corresponding 1 H-MRI and 129 Xe-MRI scans from 26 healthy participants (median age [range]: 11 [8-71]; 50% females) and 289 patients with pulmonary pathologies (median age [range]: 47 [6-83]; 51% females) were split into training (422 scans [88%]; 257 participants [82%]) and testing (58 scans [12%]; 58 participants [18%]) sets. FIELD STRENGTH/SEQUENCE: 1.5-T, three-dimensional (3D) spoiled gradient-recalled 1 H-MRI and 3D steady-state free-precession 129 Xe-MRI. ASSESSMENT: We developed a multimodal DL approach, integrating 129 Xe-MRI and 1 H-MRI, in a dual-channel convolutional neural network. We compared this approach to single-channel alternatives using manually edited LCEs as a benchmark. We further assessed a fully automatic DL-based framework to calculate VDPs and compared it to manually generated VDPs. STATISTICAL TESTS: Friedman tests with post hoc Bonferroni correction for multiple comparisons compared single-channel and dual-channel DL approaches using Dice similarity coefficient (DSC), average boundary Hausdorff distance (average HD), and relative error (XOR) metrics. Bland-Altman analysis and paired t-tests compared manual and DL-generated VDPs. A P value < 0.05 was considered statistically significant. RESULTS: The dual-channel approach significantly outperformed single-channel approaches, achieving a median (range) DSC, average HD, and XOR of 0.967 (0.867-0.978), 1.68 mm (37.0-0.778), and 0.066 (0.246-0.045), respectively. DL-generated VDPs were statistically indistinguishable from manually generated VDPs (P = 0.710). DATA CONCLUSION: Our dual-channel approach generated LCEs, which could be integrated with ventilated lung segmentations to produce biomarkers such as the VDP without manual intervention. EVIDENCE LEVEL: 4. TECHNICAL EFFICACY: Stage 1.

Large-scale investigation of deep learning approaches for ventilated lung segmentation using multi-nuclear hyperpolarized gas MRI.

Respiratory diseases are leading causes of mortality and morbidity worldwide. Pulmonary imaging is an essential component of the diagnosis, treatment planning, monitoring, and treatment assessment of respiratory diseases. Insights into numerous pulmonary pathologies can be gleaned from functional lung MRI techniques. These include hyperpolarized gas ventilation MRI, which enables visualization and quantification of regional lung ventilation with high spatial resolution. Segmentation of the ventilated lung is required to calculate clinically relevant biomarkers. Recent research in deep learning (DL) has shown promising results for numerous segmentation problems. Here, we evaluate several 3D convolutional neural networks to segment ventilated lung regions on hyperpolarized gas MRI scans. The dataset consists of 759 helium-3 (3He) or xenon-129 (129Xe) volumetric scans and corresponding expert segmentations from 341 healthy subjects and patients with a wide range of pathologies. We evaluated segmentation performance for several DL experimental methods via overlap, distance and error metrics and compared them to conventional segmentation methods, namely, spatial fuzzy c-means (SFCM) and K-means clustering. We observed that training on combined 3He and 129Xe MRI scans using a 3D nn-UNet outperformed other DL methods, achieving a mean ± SD Dice coefficient of 0.963 ± 0.018, average boundary Hausdorff distance of 1.505 ± 0.969 mm, Hausdorff 95th percentile of 5.754 ± 6.621 mm and relative error of 0.075 ± 0.039. Moreover, limited differences in performance were observed between 129Xe and 3He scans in the testing set. Combined training on 129Xe and 3He yielded statistically significant improvements over the conventional methods (p < 0.0001). In addition, we observed very strong correlation and agreement between DL and expert segmentations, with Pearson correlation of 0.99 (p < 0.0001) and Bland-Altman bias of - 0.8%. The DL approach evaluated provides accurate, robust and rapid segmentations of ventilated lung regions and successfully excludes non-lung regions such as the airways and artefacts. This approach is expected to eliminate the need for, or significantly reduce, subsequent time-consuming manual editing.

Deep learning in structural and functional lung image analysis.

The recent resurgence of deep learning (DL) has dramatically influenced the medical imaging field. Medical image analysis applications have been at the forefront of DL research efforts applied to multiple diseases and organs, including those of the lungs. The aims of this review are twofold: (i) to briefly overview DL theory as it relates to lung image analysis; (ii) to systematically review the DL research literature relating to the lung image analysis applications of segmentation, reconstruction, registration and synthesis. The review was conducted following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. 479 studies were initially identified from the literature search with 82 studies meeting the eligibility criteria. Segmentation was the most common lung image analysis DL application (65.9% of papers reviewed). DL has shown impressive results when applied to segmentation of the whole lung and other pulmonary structures. DL has also shown great potential for applications in image registration, reconstruction and synthesis. However, the majority of published studies have been limited to structural lung imaging with only 12.9% of reviewed studies employing functional lung imaging modalities, thus highlighting significant opportunities for further research in this field. Although the field of DL in lung image analysis is rapidly expanding, concerns over inconsistent validation and evaluation strategies, intersite generalisability, transparency of methodological detail and interpretability need to be addressed before widespread adoption in clinical lung imaging workflow.

Radical accelerated radiotherapy for non-small cell lung cancer (NSCLC): A 5-year retrospective review of two dose fractionation schedules.

BACKGROUND AND PURPOSE: Numerous fractionation regimes are used for inoperable NSCLC patients not suitable for stereotactic ablative radiotherapy. Continuous hyperfractionated accelerated radiotherapy (CHART, 54 Gy, 36 fractions over 12 days) and hypofractionated accelerated radiotherapy (55 Gy, 20 fractions over 4 weeks) are recommended UK schedules. In this single-centre retrospective analysis, we compare both fractionation schemes for patients treated at our institution from 2010 to 15. MATERIALS AND METHODS: Clinical demographic, tumour and survival data were collected alongside radiotherapy dosimetric data from the Varian Eclipse Scripting application programming interface. Differences were assessed using independent samples t-tests. Multivariate survival analysis was performed using Cox regression. RESULTS: We identified 563 eligible patients; 43% received CHART and 57% hypofractionated radiotherapy. Median age was 71 years, 56% were male, 95% PET staged with 53% WHO performance status 0-1. 30%, 14%, 50% and 6% were stage I, II, III and IV, respectively. 38% of patients underwent induction chemotherapy. 99% completed their prescribed radiotherapy treatment. Overall response rate was 50% with a 6.5% 90-day mortality rate. Median disease-free survival was 19 months, 50% recurred locally. Median overall survival was 22.5 months with 48% alive at 2 years. Multivariate analysis identified histology, stage, performance status, chemotherapy and radiotherapy response as independent predictors of survival; no significant differences between radiotherapy regimes were observed. CONCLUSION: In our centre, CHART and hypofractionated accelerated radiotherapy produce similar outcomes. Dose escalation studies are in progress to develop these schedules to match outcomes reported in concurrent chemo-radiation studies.

Comparison of CT ventilation imaging and hyperpolarised gas MRI: effects of breathing manoeuvre.

Image registration of lung CT images acquired at different inflation levels has been proposed as a surrogate method to map lung 'ventilation'. Prior to clinical use, it is important to understand how this technique compares with direct ventilation imaging modalities such as hyperpolarised gas MRI. However, variations in lung inflation level have been shown to affect regional ventilation distributions. Therefore, the aim of this study was to evaluate the impact of lung inflation levels when comparing CT ventilation imaging to ventilation from 3He-MRI. Seven asthma patients underwent breath-hold CT at total lung capacity (TLC) and functional residual capacity (FRC). 3He-MRI and a same-breath 1H-MRI were acquired at FRC+1L and TLC. Percentage ventilated volumes (%VVs) were calculated for FRC+1L and TLC 3He-MRI. TLC-CT and registered FRC-CT were used to compute a surrogate ventilation map from voxel-wise intensity differences in Hounsfield unit values, which was thresholded at the 10th and 20th percentiles. For direct comparison of CT and 3He-MRI ventilation, FRC+1L and TLC 3He-MRI were registered to TLC-CT indirectly via the corresponding same-breath 1H-MRI data. For 3He-MRI and CT ventilation comparison, Dice similarity coefficients (DSCs) between the binary segmentations were computed. The median (range) of %VVs for FRC+1L and TLC 3He-MRI were 90.5 (54.9-93.6) and 91.8 (67.8-96.2), respectively (p  = 0.018). For MRI versus CT ventilation comparison, statistically significant improvements in DSCs were observed for TLC 3He MRI when compared with FRC+1L, with median (range) values of 0.93 (0.86-0.93) and 0.86 (0.68-0.92), respectively (p  = 0.017), for the 10-100th percentile and 0.87 (0.83-0.88) and 0.81 (0.66-0.87), respectively (p  = 0.027), for the 20-100th percentile. Correlation of CT ventilation imaging and hyperpolarised gas MRI is sensitive to lung inflation level. For ventilation maps derived from CT acquired at FRC and TLC, a higher correlation with gas ventilation MRI can be achieved if the MRI is acquired at TLC.

Assessment of the influence of lung inflation state on the quantitative parameters derived from hyperpolarized gas lung ventilation MRI in healthy volunteers.

In this study, the effect of lung volume on quantitative measures of lung ventilation was investigated using MRI with hyperpolarized 3He and 129Xe. Six volunteers were imaged with hyperpolarized 3He at five different lung volumes [residual volume (RV), RV + 1 liter (1L), functional residual capacity (FRC), FRC + 1L, and total lung capacity (TLC)], and three were also imaged with hyperpolarized 129Xe. Imaging at each of the lung volumes was repeated twice on the same day with corresponding 1H lung anatomical images. Percent lung ventilated volume (%VV) and variation of signal intensity [heterogeneity score (Hscore)] were evaluated. Increased ventilation heterogeneity, quantified by reduced %VV and increased Hscore, was observed at lower lung volumes with the least ventilation heterogeneity observed at TLC. For 3He MRI data, the coefficient of variation of %VV was <1.5% and <5.5% for Hscore at all lung volumes, while for 129Xe data the values were 4 and 10%, respectively. Generally, %VV generated from 129Xe images was lower than that seen from 3He images. The good repeatability of 3He %VV found here supports prior publications showing that percent lung-ventilated volume is a robust method for assessing global lung ventilation. The greater ventilation heterogeneity observed at lower lung volumes indicates that there may be partial airway closure in healthy lungs and that lung volume should be carefully considered for reliable longitudinal measurements of %VV and Hscore. The results suggest that imaging patients at different lung volumes may help to elucidate obstructive disease pathophysiology and progression. NEW & NOTEWORTHY We present repeatability data of quantitative metrics of lung function derived from hyperpolarized helium-3, xenon-129, and proton anatomical images acquired at five lung volumes in volunteers. Increased regional ventilation heterogeneity at lower lung inflation levels was observed in the lungs of healthy volunteers.

The VAMPIRE challenge: A multi-institutional validation study of CT ventilation imaging.

PURPOSE: CT ventilation imaging (CTVI) is being used to achieve functional avoidance lung cancer radiation therapy in three clinical trials (NCT02528942, NCT02308709, NCT02843568). To address the need for common CTVI validation tools, we have built the Ventilation And Medical Pulmonary Image Registration Evaluation (VAMPIRE) Dataset, and present the results of the first VAMPIRE Challenge to compare relative ventilation distributions between different CTVI algorithms and other established ventilation imaging modalities. METHODS: The VAMPIRE Dataset includes 50 pairs of 4DCT scans and corresponding clinical or experimental ventilation scans, referred to as reference ventilation images (RefVIs). The dataset includes 25 humans imaged with Galligas 4DPET/CT, 21 humans imaged with DTPA-SPECT, and 4 sheep imaged with Xenon-CT. For the VAMPIRE Challenge, 16 subjects were allocated to a training group (with RefVI provided) and 34 subjects were allocated to a validation group (with RefVI blinded). Seven research groups downloaded the Challenge dataset and uploaded CTVIs based on deformable image registration (DIR) between the 4DCT inhale/exhale phases. Participants used DIR methods broadly classified into B-splines, Free-form, Diffeomorphisms, or Biomechanical modeling, with CT ventilation metrics based on the DIR evaluation of volume change, Hounsfield Unit change, or various hybrid approaches. All CTVIs were evaluated against the corresponding RefVI using the voxel-wise Spearman coefficient rS , and Dice similarity coefficients evaluated for low function lung ( DSClow ) and high function lung ( DSChigh ). RESULTS: A total of 37 unique combinations of DIR method and CT ventilation metric were either submitted by participants directly or derived from participant-submitted DIR motion fields using the in-house software, VESPIR. The rS and DSC results reveal a high degree of inter-algorithm and intersubject variability among the validation subjects, with algorithm rankings changing by up to ten positions depending on the choice of evaluation metric. The algorithm with the highest overall cross-modality correlations used a biomechanical model-based DIR with a hybrid ventilation metric, achieving a median (range) of 0.49 (0.27-0.73) for rS , 0.52 (0.36-0.67) for DSClow , and 0.45 (0.28-0.62) for DSChigh . All other algorithms exhibited at least one negative rS value, and/or one DSC value less than 0.5. CONCLUSIONS: The VAMPIRE Challenge results demonstrate that the cross-modality correlation between CTVIs and the RefVIs varies not only with the choice of CTVI algorithm but also with the choice of RefVI modality, imaging subject, and the evaluation metric used to compare relative ventilation distributions. This variability may arise from the fact that each of the different CTVI algorithms and RefVI modalities provides a distinct physiologic measurement. Ultimately this variability, coupled with the lack of a "gold standard," highlights the ongoing importance of further validation studies before CTVI can be widely translated from academic centers to the clinic. It is hoped that the information gleaned from the VAMPIRE Challenge can help inform future validation efforts.

Spatial Comparison of CT-Based Surrogates of Lung Ventilation With Hyperpolarized Helium-3 and Xenon-129 Gas MRI in Patients Undergoing Radiation Therapy.

PURPOSE: To develop and apply an image acquisition and analysis strategy for spatial comparison of computed tomography (CT)-ventilation images with hyperpolarized gas magnetic resonance imaging (MRI). METHODS AND MATERIALS: Eleven lung cancer patients underwent xenon-129 (129Xe) and helium-3 (3He) ventilation MRI and coregistered proton (1H) anatomic MRI. Expiratory and inspiratory breath-hold CTs were used for deformable image registration and calculation of 3 CT-ventilation metrics: Hounsfield unit (CTHU), Jacobian (CTJac), and specific gas volume change (CTSGV). Inspiration CT and hyperpolarized gas ventilation MRI were registered via same-breath anatomic 1H-MRI. Voxel-wise Spearman correlation coefficients were calculated between each CT-ventilation image and its corresponding 3He-/129Xe-MRI, and for the mean values in regions of interest (ROIs) ranging from fine to coarse in-plane dimensions of 5 × 5, 10 × 10, 15 × 15, and 20 × 20, located within the lungs as defined by the same-breath 1H-MRI lung mask. Correlation of 3He and 129Xe-MRI was also assessed. RESULTS: Spatial correlation of CT-ventilation against 3He/129Xe-MRI increased with ROI size. For example, for CTHU, mean ± SD Spearman coefficients were 0.37 ± 0.19/0.33 ± 0.17 at the voxel-level and 0.52 ± 0.20/0.51 ± 0.18 for 20 × 20 ROIs, respectively. Correlations were stronger for CTHU than for CTJac or CTSGV. Correlation of 3He with 129Xe-MRI was consistently higher than either gas against CT-ventilation maps over all ROIs (P < .05). No significant differences were observed between CT-ventilation versus 3He-MRI and CT-ventilation versus 129Xe-MRI. CONCLUSION: Comparison of ventilation-related measures from CT and registered hyperpolarized gas MRI is feasible at a voxel level using a dedicated acquisition and analysis protocol. Moderate correlation between CT-ventilation and MRI exists at a regional level. Correlation between MRI and CT is significantly less than that between 3He and 129Xe-MRI, suggesting that CT-ventilation surrogate measures may not be measuring lung ventilation alone.

Comparison of 3 He and 129 Xe MRI for evaluation of lung microstructure and ventilation at 1.5T.

BACKGROUND: To support translational lung MRI research with hyperpolarized 129 Xe gas, comprehensive evaluation of derived quantitative lung function measures against established measures from 3 He MRI is required. Few comparative studies have been performed to date, only at 3T, and multisession repeatability of 129 Xe functional metrics have not been reported. PURPOSE/HYPOTHESIS: To compare hyperpolarized 129 Xe and 3 He MRI-derived quantitative metrics of lung ventilation and microstructure, and their repeatability, at 1.5T. STUDY TYPE: Retrospective. POPULATION: Fourteen healthy nonsmokers (HN), five exsmokers (ES), five patients with chronic obstructive pulmonary disease (COPD), and 16 patients with nonsmall-cell lung cancer (NSCLC). FIELD STRENGTH/SEQUENCE: 1.5T. NSCLC, COPD patients and selected HN subjects underwent 3D balanced steady-state free-precession lung ventilation MRI using both 3 He and 129 Xe. Selected HN, all ES, and COPD patients underwent 2D multislice spoiled gradient-echo diffusion-weighted lung MRI using both hyperpolarized gas nuclei. ASSESSMENT: Ventilated volume percentages (VV%) and mean apparent diffusion coefficients (ADC) were derived from imaging. COPD patients performed the whole MR protocol in four separate scan sessions to assess repeatability. Same-day pulmonary function tests were performed. STATISTICAL TESTS: Intermetric correlations: Spearman's coefficient. Intergroup/internuclei differences: analysis of variance / Wilcoxon's signed rank. Repeatability: coefficient of variation (CV), intraclass correlation (ICC) coefficient. RESULTS: A significant positive correlation between 3 He and 129 Xe VV% was observed (r = 0.860, P < 0.001). VV% was larger for 3 He than 129 Xe (P = 0.001); average bias, 8.79%. A strong correlation between mean 3 He and 129 Xe ADC was obtained (r = 0.922, P < 0.001). MR parameters exhibited good correlations with pulmonary function tests. In COPD patients, mean CV of 3 He and 129 Xe VV% was 4.08% and 13.01%, respectively, with ICC coefficients of 0.541 (P = 0.061) and 0.458 (P = 0.095). Mean 3 He and 129 Xe ADC values were highly repeatable (mean CV: 2.98%, 2.77%, respectively; ICC: 0.995, P < 0.001; 0.936, P < 0.001). DATA CONCLUSION: 129 Xe lung MRI provides near-equivalent information to 3 He for quantitative lung ventilation and microstructural MRI at 1.5T. LEVEL OF EVIDENCE: 3 Technical Efficacy Stage 2 J. Magn. Reson. Imaging 2018.

Impact of field number and beam angle on functional image-guided lung cancer radiotherapy planning.

To investigate the effect of beam angles and field number on functionally-guided intensity modulated radiotherapy (IMRT) normal lung avoidance treatment plans that incorporate hyperpolarised helium-3 magnetic resonance imaging (3He MRI) ventilation data. Eight non-small cell lung cancer patients had pre-treatment 3He MRI that was registered to inspiration breath-hold radiotherapy planning computed tomography. IMRT plans that minimised the volume of total lung receiving ⩾20 Gy (V20) were compared with plans that minimised 3He MRI defined functional lung receiving ⩾20 Gy (fV20). Coplanar IMRT plans using 5-field manually optimised beam angles and 9-field equidistant plans were also evaluated. For each pair of plans, the Wilcoxon signed ranks test was used to compare fV20 and the percentage of planning target volume (PTV) receiving 90% of the prescription dose (PTV90). Incorporation of 3He MRI led to median reductions in fV20 of 1.3% (range: 0.2-9.3%; p = 0.04) and 0.2% (range: 0 to 4.1%; p = 0.012) for 5- and 9-field arrangements, respectively. There was no clinically significant difference in target coverage. Functionally-guided IMRT plans incorporating hyperpolarised 3He MRI information can reduce the dose received by ventilated lung without comprising PTV coverage. The effect was greater for optimised beam angles rather than uniformly spaced fields.

Comparison of CT-based Lobar Ventilation with 3He MR Imaging Ventilation Measurements.

PURPOSE: To compare lobar lung ventilation computed from expiratory and inspiratory computed tomographic (CT) data with direct measurements of ventilation at hyperpolarized helium 3 ((3)He) magnetic resonance (MR) imaging by using same-breath hydrogen 1 ((1)H) MR imaging examinations to coregister the multimodality images. MATERIALS AND METHODS: The study was approved by the national research ethics committee, and written patient consent was obtained. Thirty patients with asthma underwent breath-hold CT at total lung capacity and functional residual capacity. (3)He and (1)H MR images were acquired during the same breath hold at a lung volume of functional residual capacity plus 1 L. Lobar segmentations delineated by major fissures on both CT scans were used to calculate the percentage of ventilation per lobe from the change in inspiratory and expiratory lobar volumes. CT-based ventilation was compared with (3)He MR imaging ventilation by using diffeomorphic image registration of (1)H MR imaging to CT, which enabled indirect registration of (3)He MR imaging to CT. Statistical analysis was performed by using the Wilcoxon signed-rank test, Pearson correlation coefficient, and Bland-Altman analysis. RESULTS: The mean ± standard deviation absolute difference between the CT and (3)He MR imaging percentage of ventilation volume in all lobes was 4.0% (right upper and right middle lobes, 5.4% ± 3.3; right lower lobe, 3.7% ± 3.9; left upper lobe, 2.8% ± 2.7; left lower lobe, 3.9% ± 2.6; Wilcoxon signed-rank test, P < .05). The Pearson correlation coefficient between the two techniques in all lobes was 0.65 (P < .001). Greater percentage of ventilation was seen in the upper lobes with (3)He MR imaging and in the lower lobes with CT. This was confirmed with Bland-Altman analysis, with 95% limits of agreement for right upper and middle lobes, -2.4, 12.7; right lower lobe, -11.7, 4.6; left upper lobe, -4.9, 8.7; and left lower lobe, -9.8, 2.8. CONCLUSION: The percentage of regional ventilation per lobe calculated at CT was comparable to a direct measurement of lung ventilation at hyperpolarized (3)He MR imaging. This work provides evidence for the validity of the CT model, and same-breath (1)H MR imaging enables regional interpretation of (3)He ventilation MR imaging on the underlying lung anatomy at thin-section CT.

A method for quantitative analysis of regional lung ventilation using deformable image registration of CT and hybrid hyperpolarized gas/1H MRI.

Hyperpolarized gas magnetic resonance imaging (MRI) generates highly detailed maps of lung ventilation and physiological function while CT provides corresponding anatomical and structural information. Fusion of such complementary images enables quantitative analysis of pulmonary structure-function. However, direct image registration of hyperpolarized gas MRI to CT is problematic, particularly in lungs whose boundaries are difficult to delineate due to ventilation heterogeneity. This study presents a novel indirect method of registering hyperpolarized gas MRI to CT utilizing (1)H-structural MR images that are acquired in the same breath-hold as the gas MRI. The feasibility of using this technique for regional quantification of ventilation of specific pulmonary structures is demonstrated for the lobes.The direct and indirect methods of hyperpolarized gas MRI to CT image registration were compared using lung images from 15 asthma patients. Both affine and diffeomorphic image transformations were implemented. Registration accuracy was evaluated using the target registration error (TRE) of anatomical landmarks identified on (1)H MRI and CT. The Wilcoxon signed-rank test was used to test statistical significance.For the affine transformation, the indirect method of image registration was significantly more accurate than the direct method (TRE = 14.7 ± 3.2 versus 19.6 ± 12.7 mm, p = 0.036). Using a deformable transformation, the indirect method was also more accurate than the direct method (TRE = 13.5 ± 3.3 versus 20.4 ± 12.8 mm, p = 0.006).Accurate image registration is critical for quantification of regional lung ventilation with hyperpolarized gas MRI within the anatomy delineated by CT. Automatic deformable image registration of hyperpolarized gas MRI to CT via same breath-hold (1)H MRI is more accurate than direct registration. Potential applications include improved multi-modality image fusion, functionally weighted radiotherapy planning, and quantification of lobar ventilation in obstructive airways disease.

Dosimetric evaluation of inspiration and expiration breath-hold for intensity-modulated radiotherapy planning of non-small cell lung cancer.

The purpose of this study was to compare target coverage and lung tissue sparing between inspiration and expiration breath-hold intensity-modulated radiotherapy (IMRT) plans for patients with non-small cell lung cancer (NSCLC). In a prospective study, seven NSCLC patients gave written consent to undergo both moderate deep inspiration and end-expiration breath-hold computed tomography (CT), which were used to generate five-field IMRT plans. Dose was calculated with a scatter and an inhomogeneity correction algorithm. The percentage of the planning target volume (PTV) receiving 90% of the prescription dose (PTV(90)), the volume of total lung receiving >or=10 Gy (V(10)) and >or=20 Gy (V(20)) and the mean lung dose (MLD) were compared by the Student's paired t-test. Compared with the expiration plans, the mean +/- SD reductions for V(10), V(20) and MLD on the inspiration plans were 4.0 +/- 3.7% (p = 0.031), 2.5 +/- 2.3% (p = 0.028) and 1.1 +/- 0.7 Gy (p = 0.007), respectively. Conversely, a mean difference of 1.1 +/- 1.1% (p = 0.044) in PTV(90) was demonstrated in favour of expiration. When using IMRT, inspiration breath-hold can reduce the dose to normal lung tissue while expiration breath-hold can improve the target coverage. The improved lung sparing at inspiration may outweigh the modest improvements in target coverage at expiration.